ochafik/llama.cpp

Pull Requests Commits

server.py: raise n_predict

ochafik committed 2 years ago

1c3d6ae4

server.py: reenable grammar, accommodate mistral's escaped underscores

ochafik committed 2 years ago

f554b98f

json: not_literal

ochafik committed 2 years ago

4548ac7a

server.py: kinda api-compliant output, disabled grammar

ochafik committed 2 years ago

e9dbabd6

server.py: make tools work w/ mixtral-8x7b-instruct

ochafik committed 2 years ago

1f372940

server.py: better test for args in server.cpp

ochafik committed 2 years ago

9abeeee5

server.py: default tools work!

ochafik committed 2 years ago

bd9e3b83

agents: scripts to run scripts as sandboxed fastapi servers

ochafik committed 2 years ago

e007b02d

server.py: hacky code

ochafik committed 2 years ago

05cbe9e9

gguf: add ReaderField.read() method + update reader example

ochafik committed 2 years ago

768066de

json: rename python schema converter w/ underscores

ochafik committed 2 years ago

bc3a692d

json: add not_literal helper to python schema converter

ochafik committed 2 years ago

03a1c8bb

nix: fix blas support (#6281)

ck3d committed 2 years ago

Verified b06c16ef

tests : include IQ2_XXS and IQ2_XS in test-quantize-fns (#6303)

ikawrakow committed 2 years ago

Verified 1f2fd4e7

flake.lock: Update (#6266)

ggerganov committed 2 years ago

Verified 43139cc5

cuda : fix LLAMA_CUDA_F16 build (#6298)

slaren committed 2 years ago

Verified 2f34b865

cuda : refactor into multiple files (#6269)

slaren committed 2 years ago

Verified ae1f211c

Server: clean up OAI params parsing function (#6284)

ngxson committed 2 years ago

Verified ad3a0505

[SYCL] fix SYCL backend build on windows is break by LOG() error (#6290)

NeoZhangJianyu committed 2 years ago

Verified 95ad616c

examples : add "retrieval" (#6193)

mscheong01 committed 2 years ago

Verified 64e7b47c

ggml : support AVX512VNNI (#6280)

jart committed 2 years ago

Verified 7733f0c7

Fix heap corruption from wmode out-of-bound writes on windows (#6272)

TheFlipbook committed 2 years ago

Verified a32b77c4

imatrix : fix wname for mul_mat_id ops (#6271)

ggerganov committed 2 years ago

Verified a0e584de

Fixed lookup compilation issues on Windows (#6273)

JohannesGaessler committed 2 years ago

Verified 7aed0ffe

ci : close inactive issue, increase operations per run (#6270)

phymbert committed 2 years ago

Verified ea279d56

sampling : deduplicated code for probability distribution access (#6240)

mscheong01 committed 2 years ago

Verified 586e7bc5

[SYCL] offload op (#6217)

airMeng committed 2 years ago

Verified ddf65685

Support build win release for SYCL (#6241)

NeoZhangJianyu committed 2 years ago

Verified d03224ac

use _wfopen instead of fopen on Windows (#6248)

cebtenzzre committed 2 years ago

Verified 94d1b3b4

gitignore : gguf-split

ggerganov committed 2 years ago

Verified 95562175

Newer Older